Balance, Challenge & Antagonism when Running Dungeons & Dragons

D&D Cartoon Dungeon Master - pretty friendly

CHALLENGE RATING AND D&D AS A CONTEST:

The concept of Challenge Rating in D&D, a means of balancing encounters can detract from open world, location based, sandbox play - even if it is useful for tactical combat focused games.

The 5th Edition of Dungeons and Dragons has a system I find strange coming from playing earlier editions almost exclusively - the Challenge Rating ("CR").

It may seem odd, but I've never run a game where I tried to determine numerically if a combat encounter would be too dangerous for my players - this isn't to say that adventure design shouldn't require some consideration of enemy strength compared to that of the party, but really it's not an issue that older editions obsess too much on, even if it is treated as an absolutely core system to newer editions of Dungeons & Dragons. I fear that the reliance on a mechanical system to assure "balance" and "fairness" both shows a corrosive distrust of the Game Master and encourages a less open more competitive style of play.

It appears Challenge Rating started as a concept somewhere in the mass of publications that was D&D's 3.5 edition, but in 4th edition it really became a key component of Encounter Design, which itself assumed a place of prominence rather then being a small subset of Adventure Design.  It's also easy to see why Challenge Rating in 4th Edition Dungeons & Dragons, and 3.5's offshoot Pathfinder, became such a profoundly important, and useful, element of running the game.  3.5 and 4th Edition are mechanically complex games where the locus of play (a term I use frequently to describe the type of play where the players are most engaged and which takes up the most time at the table) is tactical combat.  Combat encounter design becomes the focus for most GMs in these editions and games, as it is designed to be, and CR functions as a useful set of tools to provide the balance that the highly tuned, but straight-forward, combat mechanics require to make the combat encounters both challenging and potentially survivable.  4th Edition CR rules create "XP budgets" to build and modify encounters from a selected list of pre-designed (or built and modified via templates) foes, each with a combat role that adds complexity to combat encounters and defines their tactics to the GM. 

While it's popular among some who play older editions and styles of D&D to wring hands or complain about this style of play and the need for or usefulness of a CR based combat encounter construction system in 4E and Pathfinder, it's worth noting that these GM facing systems make sense for the game being played - a complex tactical combat system about direct grid based confrontation with fantastical creatures. While, no matter the system, a good GM or designer will always want to have some idea about how an encounter or string of encounters might deplete player resources, the more complex and mechanically defined those resources are, the more systems for checking if an encounter or adventure is properly challenging for players becomes useful - especially when the goal of a play session is combat, sidestepping other resource draining activities (puzzles, negotiation) as quickly as possible to assure adequate time for complex combats.  As the game itself puts it:

"[T]he D&D game is a series of encounters. Encounters are where the game happens—where the capabilities of the characters are put to the test and success or failure hang in the balance." 4th Edition Dungeon Master's Guide Deluxe. (2008, James Wyatt) pg 34.
 
It becomes clear that combat centered gameplay within the matrix of scene based adventure design can be a choice, and the designers of D&D's 3.5 and 4th editions chose to build a system whose mechanics supported this style of play.  Yet it's equally worth noting that in doing so they chose to remove or sideline other aspects of the game that made up a large amount of play in earlier editions.  This transformation of the game also has effects that go beyond mere mechanics and change the duties and conception of what a good Game Master should do and bring to the table - and they do so in a way that has the potential to be disastrous for games where the locus of play isn't tactical combat.

Why is Challenge Rating in 5th Edition:

I can't see a good reason for Challenge Rating's persistence into 5th Edition. While the game has some tactical combat elements it appears to be intended as a return to exploration and role play.  CR may be useful for complex tactical battles as provided in 4th edition and Pathfinder, but the 5th Edition version does little to 'balance' the game and seems cursory and perhaps ineffectual for those who do wish to use it.  The negative effects of CR on game culture persist however and it can even promote antagonistic play.

 

Challenge Rating as a GM aide persists in 5th Edition which, while retaining a few complexities of the detailed tactical systems of 4E and Pathfinder, is a simpler game with a lot more space for non-combat challenges. Furthermore, the Fifth Edition Dungeon Master Guide offers little rationale for retaining the CR, giving no explanatory reasoning for its XP buy system that builds combat encounters.  As much as 4th Edition may represent a refinement of the design principles of character building and tactical combat as the primary locus of play, cutting away or streamlining legacy systems, 5th Edition appears to be an effort to return to more holistic gameplay with exploration, role play and fictional positioning based problem solving returning as potentially key elements of play.  It's possible, even likely, that the 5th Editions continued reliance on CR is an (unexamined perhaps?) holdover from 3.5 and 4th edition, the design space where its designers are most comfortable. However, the marginal benefit of a GM tool that claims to create balanced combat encounters is outweighed by it's potential downsides, both mechanically and socially in a system where exploration is a major part of gameplay and where the CR system is cursory. 

It is possible to play 5th Edition as a tactical combat game, and Challenge Rating may be quite useful for GMs who still find their primary work is to design direct confrontation combat encounters that will provide some, but not too much, challenge to players. However, for GMs who are trying to run a more classic sort of game, where combat is rarely straightforward confrontation and exploration of adventure locales or faction conflict and associated role play are the loci of play, CR is at best nearly useless as a metric, though presumably it may still be a comforting way to roughly check if a group of monsters is far too powerful for a party.   Unfortunately, beyond expending GM time, CR also creates other difficulties, not necessarily as a mechanic alone, but because it's presented as the way to provide fairness and balance.

Side Effect of CR and Competitive Play:

Challenge Rating, "Rules as Written" and Competitive Play through Adventurer's League are part of a laudable effort to encourage a minimum standard of Game Mastering and allow players to have fun independent of GM ability, but they come at a cost.  That cost is a distrust between players and Game Master and the loss of the idea that D&D is a cooperative game where the GM's power over the world comes with duties to adjudicate fairly and openly. Systems like "CR" instead encourage GM's to view the game competitively and antagonistically, which may be desirable for a war-game but is ruinous for anything more open ended and may not even be effective.

Challenge Rating is itself a harmless and somewhat questionable metric for determining difficultly, but when it's promoted as the way to determine fairness or assure an appropriate level of risk, as if fairness is the goal of Adventure Design, it becomes a stand in for distrust of the Game Master.  When players depend on encounters built with an XP based point system to have fun and avoid summary character death, there's something far deeper wrong then a mere lack of "balance".  In an open world game, CR and a concern for mechanically assuring fairness indicates a profound lack of trust in the GM, and the assumption that he will run a bad adventure if allowed to freely use his imagination.

Another element that unintentionally buttresses this fear of free or fair play, is WotC's efforts to promote a system of public play that is semi-competitive - the Adventurers League.  The promise is that these events will have a certain consistency provided and be free of awful experiences: players who show up with a munchkin kitted out with an arsenal of magic weapons that they use to bully everyone else, or a GM that facilitates some nightmare scenario of humiliation and insult.  Stories of this sort of experience are common, and I won't dismiss them.  However, to reach it's admirable goal of improving play experience WotC has intelligently limited material to certain of its published adventures and toned down some of the competitive elements in the public play system.  Out of this concern certain design trade offs were made. Organized play and a concern that if things aren't kept clear and well defined in published adventures GMs will turn abusive may be a reason why WotC's published products contain such an obsessive focus on using Difficulty Checks for most non-combat actions that have some in game importance, are overwritten and often railroad player choice. Under this system of design, even the worst GM or most pushy, awful player can't easily go against clear rules that say "X must be rolled in situation Y" to move the adventure along. 

Even if largely mechanized play, and fetishization of CRs are successful efforts to limit bad GMs and prevent them from darting off into the weeds of sadistic joy, they create a culture of second guessing and stifle GM creativity.  Players second guess their GMs and GMs second guess themselves, having built a game culture where hollering for the use of the Rules As Written (RAW) is acceptable, even when those rules are used for exploitative meta-gaming purposes or fail to grasp a situation that is occurring at the table.

With these strange efforts to curtail GM power and creativity, second guessing the person who is running an adventure for fun, WotC and CR may have a point, but the price is too high.  At it's core this solution to bad GMing is to push D&D away from what makes tabletop roleplaying different from other tabletop games (war-games and board games) or computer RPGs - the presence of an adaptable and creative person who is taking the role of facilitating play and providing the game world's response to player decision.  Distrust that the Game Master will live up to her role as a fair adjudicator of events and set aside her own story goals, fear that she will have biases for or against specific players, or a sense that the GM needs to be the most clever person at the table are depressing and paltry reasons for restricting the play of players and Game Masters who aren't terrible. 

Beyond limitation of creative play, there is also the question of efficacy. Does limiting the GMs options, promoting player power through the idea that RAW rules are correct over GM ad hoc rulings, and encouraging combat based tactical play really prevent terrible Game Masters from being terrible? 

Late 1E DMG Cover - a surly GM this one

I'm generally for mechanical and rules based solutions to game problems, but the problem of bad GMing isn't simply a question of perverse incentives or excess complexity - it's a skill and culture issue.  The Pathfinder, 4th Edition and 5th Edition rulebooks offer note about limiting antagonism between players and GMs - but this seems insufficient where the rules themselves promote distrust. It's also easy to find the complaints of 5th edition players and GMs about bad play experiences, which suggest that even if CR and GM limiting innovations are helpful, they haven't stopped the problem of antagonistic GMing.  Worse, while these sorts of mechanical brakes on GM overreach will do little or nothing to stop truly malicious Game Masters or players who wish to bend the rules via house rules, bad adjudication of the inevitable areas that are outside WotC's expectations, or a focus on mastering and meta-gaming the rules for mechanical advantage and against the spirit of play, they are often presented as a substitute for any sort of theory or suggestion on the easy to learn skills of being a good GM.

Excuses for Antagonistic GMing:

 

Antagonistic GMing has an appeal to some, and perhaps even a certain 'rightness' about it for running risky combat encounters, and unfortunately Dungeons & Dragons has promoted it to lesser or greater degrees for a long time. At its core though antagonistic GMing is destructive in games that depend on fictional positioning - that is to say include cooperative world-building, non-combat problems and non-combat solutions. 5th Edition D&D can be such a game.

 

I doubt it was Dragonlance that started encouraging trends in bad GMing, but the 'Prologue" of DL-1 Dragons of Despair is a wonderful example of promoting a basic misunderstanding of the GM's duties and position at the table.  It begins by turning the GM into a character in the game's story.

"A massive chair floats amid soft white light. The gold of its delicately carved frame gleams warmly. To the right of the chair, an ornate stand strains beneath a huge crystal globe. To the left of the chair, another stand holds a large book. Between the stands, an old man in brilliant white robes nods on a glistening throne. One hand rests on the globe, while the other lies poised on the book. The face looks as though the eyes closed only a moment ago. Yet the man does not move, nor does the thoughtful expression change. For this is Astinus of Palanthus, Lorekeeper of the World … You, Dungeonmaster, are the spirit of Astinus. You look upon your mortal body and again bid it farewell. For the greatest age in the history of this world called Krynn is about to unfold. You note its passage, walking unnoticed among the greatest of heroes, seeing history through the eyes of men and creatures, good and evil, feeling what they feel." - Dragons of Despair DL1 - (1984, Tracy Hickman) pg. 1

While this GM as God of History conceit might be cute, or even fun for some, once the Game Master is no longer a referee, an adjudicator or interpreter of the game world, and is instead something like a super powerful player - part of the action and story.  In 1984 this is a novel presentation of the GM's role within the game, but I'd suggest it's a jump in the wrong direction - the direction that, by encouraging the Game Master to think of himself as an active participant in story and game, rather then a facilitator standing outside of it, must resolve the GM's new role by demanding rules that themselves make value judgments, set difficulty, and resolve contradictions in mechanics or fictional positioning.  The rules, rather then the GM's understanding of their position as host and adjudicator, become the players' protection from frustrating types of challenge and ultimately the antagonistic GM emerges as an acceptable figure because he works within the constraints of the Rules as Written to provide sufficient challenge and play his 'character' . 

With a tactical fighting focused game (as D&D seems to have been for a decade or two) antagonistic GMing becomes more acceptable, and even laudable.  Much as the fun in playing Warhammer or another wargame is the challenge of going up against a human opponent and figuring out how win with a strategy that works within complex and rigid rules and equalized, point based forces when the locus of play is tactical combat the Game Master, as long as he has a decent system for determining difficulty, is free to have enemies fight as craftily and hard as possible.  A GM with this sort of rule-set can, and likely should, be trying to "win", so in this context a concern about antagonistic GMing is  warranted, but only as a matter of balance.  In a game with more options, where the rules can only offer suggestions for a wide variety of scenarios these systems are insufficient and conception of the GM as foe is disastrous.  Because of the variety of situations and the prominence puzzles or social conflict (factions) take in open world/location based or sandbox games they tend to set the rules as written aside in favor of fictional positioning.  Fictional Positioning is a focus on what the characters are doing within the game, rather then what the players are doing with the rules - a distinction between running a game where the way to check for secret doors is to say "My character taps the wall looking for hollow spaces that might conceal a door" and one where the player says "I have a 'Spot Hidden' skill of 12 and roll an adjusted 18 to find things in the room".  Without the skill metrics a GM can't maintain an antagonistic attitude to the players, the GM's duty isn't to stymie player success through crafty enemies and a dangerous environment, but to fairly adjudicate the players efforts solve puzzles in the game world.  This isn't to say in our example that our first GM doesn't have a variety of choices:

1. "The tapping reveals a hollow space behind the wall."
2. "The walls are thick and solid, you can't tell anything from tapping."
3. The GM decides they don't know if the tapping will work and rolls a check of some kind to determine if the search succeeds - perhaps adding a bonus or penalty based on the characters' methods.

However, the GM's choices are determined largely by his understanding of the fictional space, rather then through a close reading of rules (no chart of bonuses for searching based on wall material, thickness, and craftsmanship is needed - and there will be no point at which a player's "Spot Hidden" skill becomes automatically successful due to a lack of such a chart).  Within this system the players must trust that the GM is being fair to them and making interpretations of the game world and event that allow them a chance of success.

While 5th Edition often appears to reject fictional positioning in favor of "RAW" mechanics, antagonistic GMing and point balanced tactical combat based challenge was not always the way to play D&D, and with 5th Edition the mechanics are loose enough that the challenge of tactical combat can take a back-seat to exploration, puzzle solving and role play.  To someone versed only in the last ten or fifteen years of Dungeons & Dragons, and more familiar with computer RPGs or MMORPGs then tabletop play this may seem daunting, even impossible - but it's not hard to learn the skills of being a fair Game Master and determining success and failure without reliance on specific rules that directly apply to the current situation. 

Old School Approaches:

In earlier editions, up until the 1990's D&D didn't depend on a system for encounter design, and balance was less of a concern then fidelity to the setting fiction and a strained sort of naturalism.  This is not a difficult way to design adventures and can be learned easily.

 

Dungeons & Dragons was for the first twenty years largely an exploration game, and for all of its own complexity AD&D did not privilege tactical combat as a play goal* The 1st Edition Dungeon Master's Guide also fails to make any suggestions about monster placement or design that prefigure the encounter construction implicit in Challenge Rating.  Still, the 1st Edition does concern itself briefly and tangentially with "balance", often the ultimate goal or problem that 'CR' is said to overcome, but its methods are different and it's assumptions about the Game Master are different:

"Likewise, fit monsters and magic so as to be reasonable within the scope of your milieu and the particular facet of it concerned. Alter creatures freely, remembering balance. Hit dice, armor class, attacks and damage, magical and psionic powers are all mutable; and after players become used to the standard types a few ringers will make them a bit less sure of things. 

Devising a few creatures unique to your world is also recommended. As a DM you are capable of doing a proper job of it provided you have had some hours of hard experience with rapacious players. Then you will know not to design pushovers and can resist the temptation to develop the perfect player character killer!"  -  1st Edition Dungeon Master's Guide. (1979, Gary Gygax) pg 91.

This statement caps the section on "Monster Placement", one largely concerned with a sort of organic monster ecology (what lives where and how?) rather then it is with risk and reward - it cares about the fictional positioning of the monsters in the world over their mechanical power.  The advice above is all Gygax seems to offer on encounter design that's outside of concerns of Gygaxian naturalism, and I think this is telling.  There's a brief exhortation that the extensive random encounter tables (Appendix C) can be used to create a selection of foes balanced by level.  Of course the AD&D table in some ways mirror those of OD&D and allow for extremely dangerous creatures at even low levels - there are for example Ogres and Dragon hatchlings as rare (1 in 20) encounters for dungeon level 1.  However, the table based approach is mocked a bit by Gygax, though perhaps undeservedly given his own adventure design proclivities, as risking a "Disneyland" sort of campaign (which one assumes is the campaign equivalent of a "fun house" dungeon).  It would be quite possible to write an entire piece on the nature of Gygaxian or organic monster placement and how it distributes risk via random encounters, monster populations and faction balance - but that's beyond the scope of this discussion.  

The conflict or split between early and later editions is worth noting, because it suggests that balance (as much as it's important in early editions) is easy enough to obtain without reliance on Challenge Rating or any similarly complex, codified systems of encounter design and monster placement.  This may be because early editions assume (though in only a few years the Hickmans would already be thinking about narrative and scene based adventure design) location based game that doesn't focus on tactical combat.  I'd also suggest that the power curve within AD&D is flatter then that in 4e or 5e, with level or mechanical potency less important then luck and the ability to find advantages through fictional positioning as a means of deciding victory or defeat.

The 5E DMG Cover - I think these fellows are getting more antagonistic

These factors may motivate a lesser concern with balance, as may Gygax's own interest in a more naturalistic game world, but a different concept of balance, one that allows for frequent character death is also in play.  Closely attached to that concept of balance is a different set of assumptions about the duty and abilities of the Game Master.  Hack and Slash recently suggested that an open world/location based, traditional playstyle requires a GM with strong mental facilities or practice.  While I enjoy Courtney's general thrust that GMing is akin to shamanistic storytelling, I think it's something that can be picked up fairly easily by most people if they want to, and if they have the tools.  One such tool for running location based games is an understanding of how not to be antagonistic towards the players.  Antagonistic GMing is a shibboleth even in pathfinder and 4e, both of which contain advice that suggests it's a bad thing (though as I suggest above the rules may have good reason to work in an opposite direction) when a GM sees themselves in competition with the players, but I believe there's more to it then simply not actively seeking to do in the characters for one's sadistic joy. 

ANTAGONISTIC GAME MASTERS

First DMG Cover - The GM was always a threat!

The classic joke about Antagonistic Game Master is a anecdote about a GM who when confronted by a player about something, the more minor the quibble the better the story, declares "Rocks fall you all die!"  This of course would be a pretty nasty thing to do as a GM, really a betrayal of the players' trust and a fundamental rejection of the compact between GM and players.  Of course it's not especially helpful to know that Antagonistic GMs are those that offhandedly create an instant TPK out of pique. Most antagonistic GMing isn't going to take this extreme form, and it's harder to avoid mildly antagonistic actions like forcing players into direct confrontation regardless of their efforts to avoid it, especially for GMs who have become used to mechanical systems that elevate competitive tactical combat as the only viable play activity.  I'm not even going to focus on what makes up antagonistic GMing, because rather then a long list of "don'ts", the basics of avoiding antagonistic GMing (at least in the forms I've commonly seen it) are proactive and aren't hard.

 

Acknowledging the GM's Role and Power:

 

In an open world game a great deal of trust is given to the Game Master by the players as situations frequently arise where the GM must adjudicate complex situations these decisions, based more on setting fiction then rules.  The need for ad hoc decisions is ripe for misuse and give the GM a great deal of power, but because of this power a GM has a duty to act fairly.

The first step, as usual for these kinds of things, is to acknowledge the situation as it is.  The game provides the Game Master with power over the player characters because the Game Master controls the world and its reaction or response to player decision. The further one's setting and game world moves into sandbox/open world play and away from scripted narrative/scenes the more is this is true, the more room there is for a GM to respond poorly to unplanned and sudden player actions, and the more likely players (who expect freedom to choose there characters' actions and fates) will object to such heavy handed GMing.

It is important to recognize that GM power increases as scenarios and game events move further away from the simple and mechanically fenced in direct tactical combat scenarios (again it's not that these are bad, just not the basis of exploration games) and into a play style where challenges contain fewer dice checks and more player decisions: roleplaying negotiation with potential enemies and allies, solving puzzles, planning ambushes or other strategies to defeat powerful foes, and disarming or bypassing complex traps.  In all of these cases, while a dice roll or two may be important most results and interactions are going to be within the control of the GM and largely without specific written rules.  While reaction rolls and a charisma bonus or spell effect can add a layer of mechanical rules to a negotiation between player characters and monsters, it's still the GM's duty to decide the many things that shape the encounter and make it exciting: can the foes be reasoned with at all (your persuasion skills shouldn't work on rot grubs), what do the monsters want, what are their personalities, are they plotting a double-cross are are they honorable?  All of these considerations will make a negotiation between player characters and a monster or group of monsters fun, but they are without written rules (or almost without them) to limit or direct GM decision making.

Even where rules apply to complex issues of fiction positioning, the rules alone rarely suffice to properly adjudicate the scenario - can a player swing on a chandelier to cross a ballroom and avoid the trapped floor tiles? While a chandelier swinging action may require the player to make some sort of dexterity or acrobatics check, there's no rules based advice on exactly how far this will carry the character or what happens should the character fail.  The GM will need to decide what happens based on the situation (the setting fiction) and the player's stated action (fictive positioning) before even deciding what sort of check may be required to succeed and what failure means.  A GM could provide a very difficult check, or worse decide that success or failure will only propel the character part way over the trapped tiles, or even that the chandelier won't hold the character's weight - all of these would be correct interpretations (as would ones that made the task very easy or automatic) of the space left in the rules.  The difficulty is which choice should a good GM make and what choices would an antagonistic GM make?

At the core of this concern is the question of balance again, but no in a sense that's easy to design clear metrics for.  Balance between providing risk to characters with meaningful results for failure and making those risks or challenges feel "fair" to players.  The ultimate goal of this approach is to create an environment of mutual trust between players and GM and to achieve this 'balance' I propose two basic methods or philosophies about Game Mastering that seek to allow challenge and risk, while eliminating antagonistic solutions, even unintentional ones.

Description and Information as Fairness:

The Game Master is the players best and often only source of information on the game world, and because of this it's imperative on the GM to provide sufficient information to the players to make informed decisions.  Without such efforts risks will likely appear unfair and negative consequences seem the result of the GM withholding information and tricking player, making it important for GMs to check with players, even providing cautions and warnings, to make sure that they understand the stakes and situation their characters are involved in.


At its core a tabletop roleplaying game is a game of collective storytelling, with the Game Master creating a setting and personalities within that fictional space and the players interacting with it via their character's decisions.  To do make this relationship work between player and Game Master there needs to be trust on both sides, and the best way to do that from the Game master's side is to portray the world that the players can understand and adapt to and even where the players can't immediately understand events those event are clearly logical and explicable (within a magical fantasy world of course) in retrospect.  This isn't to say that tricks, traps and ambushes aren't acceptable for the non-antagonistic GM, of course not, only that these challenges shouldn't rely on the GM withholding information from the players but rather the fictional world's internally logical limitations on information. 

For example a bandit ambush can be well or badly designed - from a few men waiting in the ditch around a bend to archers hidden in trees behind blinds with melee fighters below concealed by wearing camouflage suits, and a log that falls across the road with traps in the fields on either side.  However, in both cases it's important that the GM doesn't obscure or reveal information that the players should or shouldn't know. The GM's duty to describe and inform can be somewhat limited by accepting a game world that contains extremely well worn concepts and fantasy tropes.  When many game world elements (creatures, factional relationships and even places) are already understood by the players as part of a broader sense of popular fantasy it reduces the burden on the GM to describe and offer explanation. Likewise a the "Session Zero" is another effort to provide players with sufficient information to predict and comprehend in fiction elements without depending solely on the GM's description.  I prefer a minimum of such expectations derived world building (beyond things like physics and general sociological or psychological concepts because as a GM I find creative and descriptive world building enjoyable, and as a player that non-cliched settings seem to provide more opportunities for wonder and exploration.

In a game where the environment is extremely important because the challenges are organic and puzzle based, the setting less full of well recognized elements, and because fictional positioning takes precedence it's exceptionally important to remember that the players only knowledge about the world comes from the GM.  The GM not only has the power to create challenges that are impossible or nearly so for the players, but also to deceive players about the world their characters are interacting with and the dangers they face - either intentionally or through omission.  Because of this it's important to describe everything in a new location or encounter that might me important, is interesting (both as camouflage for what's important and because it benefits the players efforts to understand and visualize the space) and isn't actively hidden. Read aloud, boxed text was an effort to do this, as were the illustration booklets provided in early TSR modules. 

Like 'CR' these descriptive aides sought to create standard/uniform play experience to both alleviate the need for good descriptive speaking skills on the part of the GM and protect players from antagonistic withholding of important details. However, read aloud and descriptive illustration aren't usually sufficient, both because read aloud is often of bad quality, becoming dull and repetitive, or acting only as a place for the adventure author to show off their imagined creative writing ability, and worse because it gives the GM a way to not have to understand the room contents - especially true when the boxed text (or drawing) doesn't include everything.  Better a description from an engaged GM that prompts player clarification and questions where it's unclear then boxed text that neither the players or GM understand, and which can always be used by a poor GM to justify excluding necessary information.  A GM playing directly from the book, reading the box text will often read quickly and only after the players have started asking questions notice important (often concealed) dangers such as traps or encounters in the location that are listed below and creating a situation where players blunder into things or the most exciting features of a location key can be completely missed. This often feels like a situation where the GM is tricking the players, springing sudden traps or encounters that they players feel their characters should have noticed.

The importance of recognizing that the GM is the players' only (or best) source of information about the game world, is that once the players accept that the GM has power and dominion over both the fiction/game-world and the mechanics, bad or lacking information is itself a form unfairness. Players that feel the need to constantly demand description and interrogate every element of the fictional space ("We check the ceilings."  "What do the floors look like?" "I throw flour in the air to see if there's drafts or any invisible objects") likely distrust their GM's willingness to provide sufficient descriptive cues or information that a space is especially dangerous or otherwise in need of a more through investigation.  Bad information (like that often provided by boxed text) creates a sense that the GM is an untrustworthy or at least recalcitrant source of information, trying to trick players into making a dangerous move. To avoid this, both as a way of encouraging trust between players and GM and as a way to prevent the tedium of over cautious exploration play GM's need to be capable of A) providing good description of locations, B) allowing and answering player questions and clarifications about description and C) checking with players to make sure they understand the descriptive elements of a situation before describing the results of an action.

The first of two these elements are are simply storytelling ability - can the GM visualize the space the game is occurring in and describe it. This is what boxed text is meant to do, but unlike a living GM, boxed text cannot respond to player questions, and because it doesn't require actual visualization or understanding of the space by the GM tends to make it more difficult for her to respond as well. Luckily these aren't hard skills to learn, they just require thinking about one's adventure space and what's in it: lighting, smell, materials, architecture - all can provide important and evocative detail. 

The last element "checking with players" is even easier but demands a GM willing to give up the sadistic joy of players' schemes suddenly going disastrously wrong - at least most of the time.  Players often fail to listen to or understand even clear description, and as a result frequently suggest that their characters engage in foolhardy actions: leaping into bottomless pits, charging into combat with unbeatable enemies and setting off traps intentionally.  Sometimes they mean to do this - frustrated or overcome with a bizarre whim, but often they simply misunderstand the situation that seems obvious to the GM.  While 'death by misadventure' should be the epitaph of every adventurer, 'death by misapprehension' feels unfair to most players. 

To avoid the unfair, unexpected and unintended demise of a beloved character it's useful for a GM to check with players and reiterate the facts of potentially dangerous situations whenever players are about to take an action.  Obviously don't limit checking to situations where death or another negative consequence is almost assured, because the GM's confirmation of character actions shouldn't signal inherent danger, but make sure that as a GM one checks in situations that have the appearance of risk or are an actual threat.  Checking is simple, the GM should just rephrase the situation and confirm that the player wants their character to act in a specific way.

For example, the GM has described a room with a well dark and deep, surrounded by a five foot lip of stone and player states "I leap into the well".  While it's entirely possible to simply start rolling falling damage as the character plummets 90' and three dungeon levels this action is worth checking.  The GM should say something like "Does your character really want to just jump into the well, you can't see the bottom, and torchlight reveals it's at least 40' deep".  When I encountered this situation it turned out that the player had misunderstood my description and though the well (not its lip) was 5' deep - he didn't want to plunge into a pit of unknown depth after all.  Had I simply rolled 9D6 damage, undoubtedly killing the character it would have felt unfair and it would have been my fault as the GM, because the player was operating under limited and false information, even if it wasn't my fault they misunderstood the description, that the well was dangerously deep and dark should have been entirely obvious to the character.  In this situation it was fairly obvious that the player's actions might be the result of a faulty understanding of the setting fiction, but other times it won't be and checking, clarifying and providing appropriate warnings, despite it sometimes taking up time, goes a long way to making sure that the players can trust that their Gm isn't trying to kill their characters, even if the setting fiction is.

Trusting Players and Characters

If balance and fairness is correlated with information accessibility the issue of challenge still arises in trying to understand how much information the GM should offer. Excessive skill tests and danger from mundane situations should be avoided, and to do so the GM is encouraged to assume that characters are skilled and competent at all aspects of their chosen careers.  Mechanical tests and their consequences are best limited to severe risks, and a GM should usually inform the player of the danger and risk offering an option to avoid or decline the test.  An easy and classic way to justify and mitigate antagonistically using skill tests is the concept of the Mythic Underworld - where such tests are limited to the actively malevolent environment of the dungeon or adventuring location.  As much as trusting the characters and treating them as competent adventurers a GM should trust players as well - allowing them to face challenges and respond to in game dangers without coddling or active efforts to assure PC triumph/survival. The mutual trust that avoiding antagonistic GMing brings can allow GMs and players both freedom to play more challenging games with high lethality and a greater focus on narrative through play and exploration.

If one accepts the idea that balance is achieved by providing sufficient information to players to recognize, overcome, avoid or decline challenges based on their perceptions of what the challenge entails this still leaves the question of how much information, description and clarification a GM should provide - either to maintain a sense of fair-play or promote enjoyment while retaining sufficient difficulty. While the ideas of descriptive fairness, checking to confirm character actions and
player understanding of the fictional space can help create balanced play, and encourages players to trust the GM - it can leave GM's wondering - "How much to reveal?"

Like the 'Rocks Fall" joke, there's also a set of jokes about the antagonistic extreme of not revealing information and the related concept of assuming character incompetence.  The jokes include the observation that a 2nd edition D&D house cat is more deadly then many 1st level PCs and jokes about would be heroes dying because they fail stat or skill checks to walk across bridges, unsheath their swords or otherwise engage in mundane activity.  At the core of this is a fear of antagonistic GMing, that the mechanics used to adjudicate dangerous situations (or the idea that a '1' is always a dangerous fumble) will be applied to mundane or unexpected activities.  To some extent this is a perverse over reliance on skill and combat mechanics, subtly encouraged by systems where the skill roll or DC check is used for a large number of activities such as climbing slippery stairs or peaceably interacting with NPCs, and especially when such checks are used as a substitute for GM adjudication and the trust associated with it.  It doesn't even require an intentionally antagonistic GM to fall into this sort of trap, just a sort of literalism, simulationist excess and misapplied logic such as: "if climbing the rickety stairs inside the lost shrine to steal the magic orb needs a DC check to avoid falling and taking 2D6 damage, climbing the stairs to the room at the inn after drinking should be the same."  Suddenly the world becomes incredibly dangerous to the characters, not simply during forays to the mythic underworld, but when they go about daily life - a death by perturbed house cat attack**, tree climbing, bridge crossing or other mundane misadventure.

Still, this is a form of antagonistic GMing, that simple 'soft' storytelling skills can prevent without resorting to mechanical enforcement or limitations.  The answer here is to embrace the idea of trusting the characters.  Trust that the characters a competent. Not just competent at simple tasks like climbing stairs, but competent and knowledgeable when it comes to the tasks of their chosen careers.  In gaming terms this means allowing characters to do most things without rolling any dice.  For example: an adventurer can throw a grapple up a cliff and then climb the rope without checking to see if she falls or the grapple breaks free unless there's something extra (arrows flying at her or a severe time crunch) that would make doing so hard for someone well practiced in the act. Likewise, as competent in wilderness survival, spelunking and similar activities characters will recognize common dangers that their players don't: dangerously unstable construction, rivers that will sweep away the unprepared or that a bear is rabid and/or has been enchanted to attack.  The idea of PC competence also helps make the game world feel more serious - because the situations and foes that the characters face are dangerous despite their being hardened and skilled adventurers.  No special feats or additional bonuses are required to empower the PCs, simply the players' trust that the GM won't inflict harm and failure from everyday or mundane tasks.  That the characters are skilled riders by default, means they don't fall off their horses and break their necks while ambling through the countryside, or even riding hard in pursuit along a rocky road.  It also insures that when the GM asks for a stat or skill check to leap a wall on a galloping horse (and again the GM should make sure to warn the player of the skilled horseman that they can either stop, change direction or make a check to leap the wall with a risk of falling)  it's not another in a series of insignificant actions - but instead a daring feat of horsemanship with a success worthy of the risk associated.  A GM should trust that

Linked to this idea, and perhaps useful in determining when a skill roll or stat check is important (though not without its own problems) is the idea of the Mythic Underworld.  In the oldest editions of D&D and its earliest campaigns the outside world and the adventuring dungeon were often clearly demarcated and dangers largely limited to events within the dungeon.  Beyond this simple haven v. adventuring location dichotomy the Mythic Underworld provides a further sense that the adventuring space is actively hostile to the characters: all doors are jammed, light attacks aggressive wandering monsters and traps only work on the party.  The Underworld itself is a malevolent intelligence seeking to injure and defeat the characters - whose prior knowledge as hardened veteran soldiers, learned sorcerers and adept adventurers allows the barest chance of survival against the power and supernatural evil of the Underworld.  Mechanically this means that while a character never slips on the inn stairs, or dies of food poisoning from bad rations - the traitorous stairs of the dungeon encourage falls, the water in its wells is always poisoned and the darkness of its corners coalesces to help the ambushes of the dungeon's monstrous denizens.  Statistic and skill checks are regularly necessary to avoid injury in such a place. Personally I like the Mythic Underworld/Mythic Wilderness as a game conceit, but am less positive towards it as a GMing instruction.  The Mythic Underworld is limiting when it comes to game options, and while it informs the players where to expect dangers, it doesn't really act to remedy antagonistic GMing, only to limit it to a certain location or environment.  Still if the idea of character competence seems too hard to apply, a strict adherence to limiting rolls and danger to the Mythic Underworld may help reign in antagonistic GMing impulses or at least make them more palatable to one's players.

On the other side of the bad GMing coin from infantilizing characters is the mirrored idea that the players themselves can't be trusted - that PCs survival or victory is an inviolable right, that a natural '20' succeeds at any action, no matter how absurd and that it's the GMs job to make sure survival and victory come to pass in the name of fin.  This is where I bring out my stereotypical old-school  grognard self and say things like "characters are meant to die".  Yet, this gruff brutality means here is that if players understand that their characters are taking risks and trust the GM not to arbitrarily introduce deadly dice tests or force them to make decisions based on bad and unreasonably withheld information there's no reason to protect their PCs overly much from negative consequences.  A reasonable player that can trace their failure and resulting character death (or whatever else your campaign dishes out as a bad end) to their own choices, clearly seeing how they could have chosen differently had they realized X or connected clue Y won't be aggrieved (sad yes, and as a GM it's okay to be sad about the death of a favorite PC) but is likely instead to feel like they should roll up someone new and push on with a bit of wisdom or a new plan.

While some players don't do this and complain about losing characters to even the most obvious mistakes or bad decisions - they're unreasonable.  The right to fairly present deadly challenges as a GM is the advantage that's bought with accepting the duty to provide sufficient information, check against misunderstandings and allow characters a wide range of actions without risk - and best of all your players will usually survive these challenges.  As much as the admonishment to 'Trust the Character' is an instruction for a GM to be forgiving and liberal with what one allows characters to accomplish without risking a dice check, a good Gm should trust the players as well.  Trust that the trust of several minds that makes up the party will puzzle through risks and dangers, thinking up ways to overcome problems the GM places before them.  Trust them to retreat, adapt, invent new tactics, unravel mysteries and think up ways through obstacles that the Gm hasn't guessed at. Also trust the players to see that your efforts at GMing are efforts at fairness and impartial adjudication rather then antagonism, and to enter into the idea that tabletop games are an exercise in collective storytelling rather then rules knowledge, pettifoggery and competition.

If you can achieve trust with your players and if they know that as a GM you're simply relaying the game fiction, describing the setting as best as you can, open to their solutions to problems and uninterested in their character's demise you may find that mechanical bulwarks against bad GMing aren't necessary and players will respect the GM's role as arbitrator in a way that allows the game to move beyond tactical combat and better encourages the exploration of the game world and freedom to present open ended/non-scripted risks and dangers to the players for them to solve.

At some point I hope to add to this piece with a discussion about how to write exploration based, non-antagonistic adventures that use an organic/narrative (as opposed to mechanics such as CR) basis for their construction and how this elevates location based adventures and faction play, but that goes beyond the scope of this article - which is already too long. 

As A GM aspire to be like this fellow
from Yawning Portal - a funny looking
bartender with some good stories.

*The discussion of combat in the 1978 Dungeon Masters Guide seems more focused on justifying the HP system, and rejecting hit locations - perhaps in response to Arneson's hit location based combat system in Blackmoor?

** In the Monstrous Compendium a domestic cat has the same AC as a warrior in scale armor, the same number of HP as a 1st level magic-user, and up to three attacks per round capable of doing a total 5HP of damage if they all hit.  It's not to say it's a nightmarish beast, but it's easily capable of doing in a 0 level human or low level adventurer if it gets lucky.  This strikes me as more of an example of 2e and the late TSR era's obsession with universal taxonomy and simulationism.  In the event you need to have house cats attack armored adventurers I'd recommend using a swarm mechanic (assuming the cats are somehow magically compelled or something).